This repository was archived by the owner on Jul 24, 2023. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 1
Expand file tree
/
Copy pathprojectionOld.html
More file actions
429 lines (387 loc) · 18.5 KB
/
projectionOld.html
File metadata and controls
429 lines (387 loc) · 18.5 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<title>Data Science Meetup</title>
<meta name="description" content="Big data technologies used at Real Impact Analytics">
<meta name="author" content="Karim Douieb">
<meta name="apple-mobile-web-app-capable" content="yes" />
<meta name="apple-mobile-web-app-status-bar-style" content="black-translucent" />
<meta name="viewport" content="width=device-width, initial-scale=1.0, maximum-scale=1.0, user-scalable=no, minimal-ui">
<link rel="stylesheet" href="css/reveal.css">
<link rel="stylesheet" href="css/theme/karim.css" id="theme">
<link rel="stylesheet" href="css/viz/viz.css">
<link rel="stylesheet" href="node_modules/font-awesome/css/font-awesome.css">
<!-- For syntax highlighting -->
<link rel="stylesheet" href="lib/css/monokai_sublime.css">
<!-- Include the appropriate print stylesheet -->
<script>
var link = document.createElement( 'link' );
link.rel = 'stylesheet';
link.type = 'text/css';
link.href = window.location.search.match( /print-pdf/gi ) ? 'css/print/pdf.css' : 'css/print/paper.css';
document.getElementsByTagName( 'head' )[0].appendChild( link );
</script>
<!--[if lt IE 9]>
<script src="lib/js/html5shiv.js"></script>
<![endif]-->
</head>
<body>
<div class="reveal">
<!-- Any section element inside of this container is displayed as a slide -->
<div class="slides">
<section>
<h1>Big Data Stack</h1>
<h3><a href="https://realimpactanalytics.com/en"><i class="fa fa-at fa-1x"></i> Real Impact Analytics</a></h3>
<h4>
<a style="color: black;" href="http://www.ulb.ac.be/di/algo/kdouieb/">Karim Douïeb</a>
</4>
</section>
<section data-state="uncenter">
<section>
<h2>Who am I?</h2>
<img style="border:none; background:none; box-shadow:none;" src="resources/learningCurve.png">
</section>
<section>
<img style="border:none; background:none; box-shadow:none;" src="resources/venn.png">
</section>
</section>
<!--<section data-state="uncenter darkTheme" data-background-iframe="https://realimpactanalytics.com/" >
<h2>Where I work</h2>
</section>-->
<section>
<section data-state="uncenterDarkTheme" data-background="resources/MapReduce/MapReduce.013.jpg">
<h2>Where do I work?</h2>
</section>
<section>
<h1>Missions</h1>
<table style="font-size: 80%;">
<thead>
<tr>
<th>Campaign managment</th>
<th>Operational optimization</th>
</tr>
</thead>
<tbody>
<tr>
<td>
Improve marketing<br>
Strengthen subscriber acquisitions<br>
Acquire competitor market<br>
Increase cross-sales<br>
Reduce churn<br>
Target high value subscribers<br>
A/B test campaign messages<br>
Compute campaign's ROI
</td>
<td>
Drive field actions<br>
Improve network coverage<br>
Performance tracking<br>
Stock management<br>
Managed services<br>
Data commercialization
</td>
</tr>
</tbody>
</table>
</section>
<section>
<h1>Data</h1>
<table>
<tbody>
<tr> <td> <i class="fa fa-phone fa-1x"></i> </td> <td> Call Detail Record (CDR) </td> </tr>
<tr> <td> <i class="fa fa-clock-o fa-1x"></i> </td> <td> Date — Time — Duration </td> </tr>
<tr> <td> <i class="fa fa-dollar fa-1x"></i> </td> <td> Billing — Recharge — Money transfer </td> </tr>
<tr> <td> <i class="fa fa-group fa-1x"></i> </td> <td> Social Network </td> </tr>
<tr> <td> <i class="fa fa-globe fa-1x"></i> </td> <td> Antenna's geolocalisation </td> </tr>
<tr> <td> <i class="fa fa-signal fa-1x"></i> </td> <td> Network information </td> </tr>
<tr> <td> <i class="fa fa-file-code-o fa-1x"></i> </td> <td> Data usage — URL — App </td> </tr>
</tbody>
</table>
</section>
<section>
<h2>Treating telco/banking data of 300,000,000 subscribers daily</h2>
</section>
<section data-state="worldTour">
<h2>Active in 25+ countries</h2>
<div class="worldTourViz"></div>
<h4 class="countryName"></h4>
</section>
<section data-state="worldRatio">
<h2>Is this a lot?</h2>
<div class="worldRatioViz"></div>
<h4 class="ratio"></h4>
<h4>4.1% of world population</h4>
</section>
<section>
<h2>Still Room for growth</h2>
</section>
</section>
<section>
<section data-background="resources/Tesla.jpg">
<h1>Spark</h1>
</section>
<section>
<div data-timeline="resources/timeline/spark.json"></div>
</section>
<section>
<h3>Map Reduce is like <span class="fragment fade">making burgers</span></h3>
<img width="700" style="border:none; background:none; box-shadow:none;" class="fragment fade" data-src="resources/Burger.jpg">
</section>
<section data-transition="fade">
<img width="800" style="border:none; background:none; box-shadow:none;" data-src="resources/MapReduce/MapReduce.001.jpg">
</section>
<section data-transition="fade">
<img width="800" style="border:none; background:none; box-shadow:none;" data-src="resources/MapReduce/MapReduce.002.jpg">
</section>
<section data-transition="fade">
<h2>MapReduce data sharing primitive</h2>
<img width="800" style="border:none; background:none; box-shadow:none;" data-src="resources/MapReduce/MapReduce.003.jpg">
</section>
<section>
<h2>Restricted programming framework</h2>
<blockquote style="width: 100%;" cite="http://www.cs.berkeley.edu/~matei/papers/2010/hotcloud_spark.pdf">
“<strong>MapReduce</strong> tasks must be written as <strong>acyclic dataflow programs</strong>, i.e. a stateless mapper followed by a stateless reducer, that are executed by a batch job scheduler.
<br>This <strong>paradigm</strong> makes repeated querying of datasets difficult and <strong>imposes limitations</strong> that are felt in fields such as machine learning, where <strong>iterative algorithms</strong> that revisit a single working set multiple times are the norm.”
<br>— <a href="http://www.cs.berkeley.edu/~matei/papers/2010/hotcloud_spark.pdf">Zaharia, Chowdhury, Franklin, Shenker, Stoica</a>
</blockquote>
</section>
<section data-transition="fade">
<img width="800" style="border:none; background:none; box-shadow:none;" data-src="resources/MapReduce/MapReduce.004.jpg">
</section>
<section data-transition="fade">
<img width="800" style="border:none; background:none; box-shadow:none;" data-src="resources/MapReduce/MapReduce.005.jpg">
</section>
<section style="font-size: 80%;">
<h2>Resilient Distributed Dataset (RDD)</h2>
<p>
<h5 style="text-align: left;">Restricted form of distributed shared memory:</h5>
<ul>
<li>Immutable, partitioned data set.</li>
<li>Can only be build through coarse-grained deterministic transformations (map, filter, join, …). These are operations that are applied to all of the elements of the data set at once.</li>
</ul>
</p>
<p>
<h5 style="text-align: left;">Efficient fault recovery using lineage:</h5>
<ul>
<li>Remember the graph of transformation applied to each RDD.</li>
<li>When is lost simply recompute lost fragment of the execution graph.</li>
</ul>
</p>
</section>
<section>
<table style="font-size: 80%;">
<thead>
<tr><th>Transformations</th><th>Actions</th></tr>
</thead>
<tbody>
<tr>
<td>
map<br>filter<br>flatMap<br>join<br>union<br>distinct<br>groupByKey<br>aggregateByKey<br>sortByKey<br>cogroup<br>sample<br>intersection<br>...
</td>
<td>
reduce<br>collect<br>count<br>first<br>take<br>takeSample<br>takeOrdered<br>countByKey<br>foreach<br>saveAsParquetFile<br>saveAsTextFile<br>saveAsObjectFile<br>...
</td>
</tr>
</tbody>
</table>
</section>
<section>
<h2>Log Mining example</h2>
<pre><code class="scala"> val file = spark.textFile("hdfs://...")
val errors = file.filter(line => line.contains("ERROR"))
errors.cache()
// Count errors mentioning MySQL
errors.filter(line => line.contains("MySQL")).count()
// Count errors mentioning Windows should be much faster
errors.filter(line => line.contains("Windows")).count()
</code></pre>
<p style="position: relative;">
<p style="position: absolute;" class="fragment"><img width="600" style="border:none; background:none; box-shadow:none;" data-src="resources/MapReduce/MapReduce.006.jpg"></p>
<p style="position: absolute;" class="fragment"><img width="600" style="border:none; background:none; box-shadow:none;" data-src="resources/MapReduce/MapReduce.007.jpg"></p>
<p style="position: absolute;" class="fragment"><img width="600" style="border:none; background:none; box-shadow:none;" data-src="resources/MapReduce/MapReduce.008.jpg"></p>
<p style="position: absolute;" class="fragment"><img width="600" style="border:none; background:none; box-shadow:none;" data-src="resources/MapReduce/MapReduce.009.jpg"></p>
<p style="position: absolute;" class="fragment"><img width="600" style="border:none; background:none; box-shadow:none;" data-src="resources/MapReduce/MapReduce.010.jpg"></p>
</p>
</section>
<section>
<h2>Logistic Regression Example</h2>
<pre><code class="scala"> val points = spark.textFile(...).map(parsePoint).cache()
var w = Vector.random(D) // current separating plane
for (i <- 1 to ITERATIONS) {
val gradient = points.map(p =>
(1 / (1 + exp(-p.y*(w dot p.x))) - 1) * p.y * p.x
).reduce(_ + _)
w -= gradient
}
println("Final separating plane: " + w)
</code></pre>
<img class="fragment fade" style="border:none; background:none; box-shadow:none;" data-src="resources/logistic-regression.png">
</section>
<section>
<h2>Spark Ecosystem</h2>
<img width="600" style="border:none; background:none; box-shadow:none;" data-src="resources/spark-stack.png">
</section>
<section>
<h2>Spark Drawback</h2>
<ul>
<li>Lack of maturity</li>
<li>Sometimes less efficient then specialized system</li>
<li>Difficult to debug
<ul>
<li>lazyness & optimisation</li>
<li>limited monitoring tools</li>
</li>
</ul>
</section>
<section>
<h2>Big Data <i class="fa fa-at fa-1x"></i> Real Impact Analytics</h2>
<p>
<i class="fa fa-phone fa-2x"></i>
<i class="fa fa-clock-o fa-2x"></i>
<i class="fa fa-dollar fa-2x"></i>
<i class="fa fa-group fa-2x"></i>
<i class="fa fa-globe fa-2x"></i>
<i class="fa fa-signal fa-2x"></i>
<i class="fa fa-file-code-o fa-2x"></i>
</p>
<p>
<i class="fa fa-arrow-circle-down fa-2x"></i>
</p>
<ul>
<li>Basic usage statistic</li>
<li>Social communities detection</li>
<li>Mobility pattern</li>
<li>Behavior prediction</li>
<li>Recharge frequency</li>
<li>Anomaly detection</li>
<li>KPIs GeoMapping</li>
</ul>
</section>
</section>
<section data-background="resources/MapReduce/MapReduce.012.jpg">
</section>
<section>
<section data-background="resources/ebolaOutbreak.png">
</section>
<section>
<h2>Leveraging anonymized telecom data to fight the Ebola outbreak</h2>
<!--<p>
collecting dynamic and updated data is one of the biggest challenges for health actors
</p>-->
</section>
<section>
<h2>4 key actors</h2>
<ul>
<li>Mobile Network Operators (provide data)</li>
<li>Health actors (use the analyses made on the data)</li>
<li>Analytics experts (make the data valuable)</li>
<li>Regulators (approve the use of the data for a specific scope and under a clearly defined approach)</li>
</ul>
</section>
<section>
<h2>Objectives</h2>
<ul>
<li>Set-up and secure a system (hardware and software) for collecting telecom and public health data</li>
<li>Analyse mobility paterns and develop epidemiological models to simulate the potential spread paths of the disease</li>
<li>Build a platform/tools allowing identification of priority locations for action</li>
</ul>
</section>
<section data-state="fullMixModel">
<h2>Epidemiological fully mixed model</h2>
<div class="fullMixModelViz"></div>
</section>
<section data-state="mobilityModel">
<h2>Hybrid Epidemiological model</h2>
<div class="mobilityModelViz"></div>
</section>
<section data
<section data-background-video="resources/Epidemics animation.webm" data-background-color="#000000">
</section>
</section>
<section>
<svg width="640" height="300" xmlns="http://www.w3.org/2000/svg">
<title>black Star Glasses</title>
<metadata id="metadata7">image/svg+xml</metadata>
<g>
<title>Layer 1</title>
<path d="m445.72406,86.045715c-37.811279,0.253937 -74.24823,12.123665 -111.362091,16.8703c-15.489655,1.981148 -31.454407,2.466522 -46.970093,0.624634c-20.629822,-2.449356 -40.603043,-8.955338 -61.219223,-11.871849c-18.424255,-2.604774 -37.078735,-5.326279 -55.618668,-5.434509c-32.400101,-0.192863 -63.735413,9.994659 -95.497093,15.495583c0,9.330353 -1.888023,21.307236 2.860909,27.865662c4.898636,6.764206 13.161667,12.950882 15.304817,22.117264c5.082314,21.740097 12.298714,43.050537 21.525597,63.48024c4.483597,9.928223 12.946198,16.849899 22.396454,21.61795c22.757065,11.481857 49.691574,13.170486 74.656403,9.559631c35.552444,-5.141983 54.840805,-31.9319 69.555923,-62.542679c5.507599,-12.442932 6.122192,-26.803879 13.376007,-38.176422c4.307831,-6.753479 14.312561,-3.910828 21.462097,-3.935516c7.150635,-0.024628 13.211548,11.919037 15.429535,19.869354c7.853668,28.134583 20.947266,57.984634 46.971222,72.725891c38.861298,22.014374 100.085052,17.896744 125.299316,-21.805481c9.867584,-15.538422 12.592499,-34.3909 17.543243,-51.982269c3.013977,-10.710403 5.707092,-23.47908 16.425049,-29.366791c4.32489,-2.376541 1.289368,-11.857956 1.927734,-17.369583c0.696289,-6.004509 2.03772,-11.316864 -6.470276,-12.495476c-14.838745,-2.058281 -29.232941,-8.315689 -44.4198,-10.433968c-14.391998,-2.007927 -28.492554,-4.909439 -43.177063,-4.811966zm-249.973877,9.996819c10.733856,-0.06752 21.421265,0.7211 31.8535,2.936897c62.158127,13.203697 63.414536,38.923218 37.826004,92.096985c-20.850906,43.326981 -64.043854,54.183075 -107.566757,43.797348c-20.063965,-4.78627 -33.782387,-18.85141 -40.252647,-38.112137c-6.829735,-20.328979 -11.994827,-41.926544 -10.825729,-63.293793c1.567093,-28.610336 33.790306,-32.631584 56.863731,-35.238449c10.587601,-1.195755 21.36911,-2.118286 32.101898,-2.186852zm226.954834,0c10.733887,0.068565 21.515411,0.991096 32.103027,2.186852c23.072235,2.606865 55.29776,6.628113 56.86261,35.238449c1.169067,21.367249 -3.996063,42.964813 -10.825714,63.293793c-6.470337,19.260727 -20.188721,33.325867 -40.251556,38.112137c-43.521759,10.385727 -86.715851,-0.470367 -107.567902,-43.797348c-25.588501,-53.173767 -24.330963,-78.893288 37.827148,-92.096985c10.432281,-2.215797 21.119659,-3.004417 31.852386,-2.936897z" stroke-width="5" stroke="#000000" fill="#000000" id="path2830"/>
</g>
</svg>
<h1>We are hiring (nerds)!</h1>
<h3>kdo@ria.bi</h3>
</section>
<section data-background-video="resources/KYC-QC.webm" data-background-color="#000000">
</section>
</div>
</div>
<script src="lib/js/head.min.js"></script>
<script src="js/reveal.js"></script>
<script src="js/d3.min.js"></script>
<script src="js/topojson.v1.min.js"></script>
<script src="js/queue.v1.min.js"></script>
<script src="node_modules/underscore/underscore.js"></script>
<!-- Visualisation -->
<script type="text/javascript" src="js/viz/worldtour/worldtour.js"></script>
<script type="text/javascript" src="js/viz/worldRatio.js"></script>
<script type="text/javascript" src="js/viz/fullMixModel.js"></script>
<script type="text/javascript" src="js/viz/mobilityModel.js"></script>
<script>
// Full list of configuration options available here:
// https://github.com/hakimel/reveal.js#configuration
Reveal.initialize({
controls: true,
progress: true,
center: true,
history: true,
touch: true,
theme: Reveal.getQueryHash().theme, // available themes are in /css/theme
transition: Reveal.getQueryHash().transition || 'default', // default/cube/page/concave/zoom/linear/fade/none
// Optional libraries used to extend on reveal.js
dependencies: [
{ src: 'lib/js/classList.js', condition: function() { return !document.body.classList; } },
{ src: 'plugin/markdown/marked.js', condition: function() { return !!document.querySelector( '[data-markdown]' ); } },
{ src: 'plugin/markdown/markdown.js', condition: function() { return !!document.querySelector( '[data-markdown]' ); } },
{ src: 'plugin/highlight/highlight.js', async: true, condition: function() { return !!document.querySelector( 'pre code' ); }, callback: function() { hljs.initHighlightingOnLoad(); } },
{ src: 'plugin/zoom-js/zoom.js', async: true },
{ src: 'plugin/notes/notes.js', async: true },
{ src: 'plugin/timeline/reveal-timeline.js', async: true, condition: function() { return 1; } }
]
});
Reveal.addEventListener( 'slidechanged', function( event ) {
Reveal.configure(currentConfig);
currentConfig = setToDefault();
} );
var setToDefault = function(){
return {center: true, theme: 'karim'};
}
var currentConfig = setToDefault();
Reveal.addEventListener( 'uncenter', function() {
currentConfig.center = false;
} );
Reveal.addEventListener( 'darkTheme', function() {
currentConfig.theme = 'karimBlack';
} );
Reveal.addEventListener( 'uncenterDarkTheme', function() {
currentConfig.center = false;
currentConfig.theme = 'karimBlack';
} );
Reveal.addEventListener( 'worldTour', function() {
worldTour();
} );
Reveal.addEventListener( 'worldRatio', function() {
worldRatio();
} );
Reveal.addEventListener( 'worldRatio', function() {
worldRatio();
} );
Reveal.addEventListener( 'fullMixModel', function() {
fullMixModel();
} );
Reveal.addEventListener( 'mobilityModel', function() {
mobilityModel();
} );
</script>
</body>
</html>